Skip to content
The Chat Archive
gpt-4-0314· OpenAI WildChat-4.8M (AI2)

写一段Python代码,将一个视频分割成几个小视频

  1. User

    写一段Python代码,将一个视频分割成几个小视频

  2. Assistant

    需要安装: pip install moviepy

    然后运行以下代码:

    from moviepy.editor import VideoFileClip
    
    def split_video(input_video, output_folder, clip_duration):
        video = VideoFileClip(input_video)
        video_duration = video.duration
        clip_number = int(video_duration / clip_duration)
    
        for i in range(clip_number):
            start_time = i * clip_duration
            end_time = (i + 1) * clip_duration
    
            # 从大视频中提取时间区间的片段并保存
            subclip = video.subclip(start_time, end_time)
            output_video = f"{output_folder}/{i}.mp4"
            subclip.write_videofile(output_video, codec="libx264", audio_codec="aac")
    
            print(f"Generated clip: {output_video}")
    
        # 处理最后一段可能的多余时间
        if video_duration % clip_duration:
            start_time = clip_number * clip_duration
            end_time = video_duration
    
            subclip = video.subclip(start_time, end_time)
            output_video = f"{output_folder}/{clip_number}.mp4"
            subclip.write_videofile(output_video, codec="libx264", audio_codec="aac")
    
            print(f"Generated clip: {output_video}")
    
        video.close()
    
    
    if __name__ == "__main__":
        input_video = "your_video.mp4"  # 视频输入路径
        output_folder = "your_output_folder"  # 输出小视频文件夹路径
        clip_duration = 10  # 要分割的小视频持续时间,单位为秒
    
        split_video(input_video, output_folder, clip_duration)
    

    使用时,请将your_video.mp4替换为您要拆分的视频的路径,并将your_output_folder替换为保存小视频的文件夹的路径。

  3. User

    写一段Python代码,将一个视频按照给的多个指定时间点分割成多个小视频

  4. Assistant

    需要安装: pip install moviepy

    然后运行以下代码:

    from moviepy.editor import VideoFileClip
    
    def split_video(input_video, output_folder, time_points):
        video = VideoFileClip(input_video)
    
        for i in range(len(time_points)-1):
            start_time = time_points[i]
            end_time = time_points[i+1]
    
            # 从大视频中提取时间区间的片段并保存
            subclip = video.subclip(start_time, end_time)
            output_video = f"{output_folder}/{i}.mp4"
            subclip.write_videofile(output_video, codec="libx264", audio_codec="aac")
    
            print(f"Generated clip: {output_video}")
    
        video.close()
    
    
    if __name__ == "__main__":
        input_video = "your_video.mp4"  # 视频输入路径
        output_folder = "your_output_folder"  # 输出小视频文件夹路径
        time_points = [0, 10, 30, 50]  # 指定时间点,单位为秒
    
        split_video(input_video, output_folder, time_points)
    

    使用时,请将your_video.mp4替换为您要拆分的视频的路径,并将your_output_folder替换为保存小视频的文件夹的路径。同时,根据需要调整time_points变量中的时间点。

    注意:确保time_points至少包含两个时间点(0秒和一个大于0的值)。将最小值保持为0秒,并保持最大值小于或等于视频持续时间。此外,按升序排列时间点,不要包含重复时间点。

  5. User

    写一段Python代码,根据视频内容生成字幕,并将字幕嵌入到视频中

  6. Assistant

    首先,我们需要使用Google Speech-to-Text API来将视频中的音频转换为文本。请在https://cloud.google.com/speech-to-text/docs/quickstart-client-libraries#client-libraries-usage-python 上查看如何在Python中设置和使用Speech-to-Text API。 确保已经安装所需的依赖:pip install --upgrade google-cloud-speech。

    然后,安装moviepy库:pip install moviepy。

    最后,运行以下代码:

    import os
    from google.cloud import speech_v1p1beta1 as speech
    from moviepy.editor import *
    
    def transcribe_audio(input_audio):
        client = speech.SpeechClient()
        audio = speech.RecognitionAudio(uri=input_audio)
    
        config = speech.RecognitionConfig(
            encoding=speech.RecognitionConfig.AudioEncoding.LINEAR16,
            language_code="en-US",
            enable_automatic_punctuation=True,
        )
    
        operation = client.long_running_recognize(config=config, audio=audio)
    
        response = operation.result(timeout=360)
    
        transcript = ""
        for result in response.results:
            transcript += result.alternatives[0].transcript
    
        return transcript
    
    def generate_subtitles(input_video, transcript):
        clips = []
        video = VideoFileClip(input_video)
        video_duration = video.duration
        text_duration = 5  # 每行字幕的持续时间,单位为秒
    
        for i in range(0, len(transcript), text_duration):
            txt_clip = TextClip(
                transcript[i:i + text_duration],
                fontsize=24,
                color="white",
                size=video.size,
                method="caption",
                align="South",
                interline=10
            ).set_duration(text_duration).set_position(("bottom"))
    
            clips.append(txt_clip)
    
        subtitles = concatenate(clips, padding=-0.1)  # 设置间隔以处理重叠的字幕
        final_video = CompositeVideoClip([video, subtitles.set_start(0).set_duration(video_duration)])
    
        return final_video
    
    def main():
        input_audio = "your_audio_file.wav"  # 音频输入路径
        input_video = "your_video.mp4"  # 视频输入路径
        output_video = "output_video.mp4"  # 输出视频路径
    
        os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = "path/to/your/credentials.json"
    
        transcript = transcribe_audio(input_audio)
        final_video = generate_subtitles(input_video, transcript)
    
        final_video.write_videofile(output_video, codec="libx264", audio_codec="aac")
        final_video.close()
    
    if __name__ == "__main__":
        main()
    

    使用时,请将your_audio_file.wav替换为您要转录的视频的音频文件路径。将your_video.mp4替换为您要添加字幕的视频的路径。如果音频和视频是独立的文件,请确保它们彼此同步。替换output_video.mp4为带有字幕的输出视频的路径。将path/to/your/credentials.json替换为Google Cloud API凭据文件的路径。

    注意:此代码示例依赖于Google Cloud API(付费服务)。还要确保将.wav音频文件与视频同步,因为Speech-to-Text API需要该格式。 若要从视频中提取音频,可以使用moviepy库的audiofileextractor模块。

  7. User

    写一段Python代码,没有声音和字幕的视频,根据该视频的内容自动生成一个字幕,并将字幕嵌入到视频中

  8. Assistant

    自动生成字幕需要两个步骤:首先从视频中提取音频文件,然后将音频转录为文字,然后将字幕嵌入到视频中。本示例将使用 moviepy 变换视频和提取音频,使用Google Speech-to-Text API将音频转录为文字。

    安装所需库:

    pip install moviepy
    pip install --upgrade google-cloud-speech
    

    将 GOOGLE_APPLICATION_CREDENTIALS 环境变量设置为 Google Cloud API 凭据的路径(JSON 文件),然后运行以下代码:

    import os
    from google.cloud import speech_v1p1beta1 as speech
    from moviepy.editor import *
    
    def transcribe_audio(input_audio):
        client = speech.SpeechClient()
        audio = speech.RecognitionAudio(uri=input_audio)
    
        config = speech.RecognitionConfig(
            encoding=speech.RecognitionConfig.AudioEncoding.LINEAR16,
            language_code="en-US",
            enable_automatic_punctuation=True,
        )
    
        operation = client.long_running_recognize(config=config, audio=audio)
    
        response = operation.result(timeout=360)
    
        transcript = ""
        for result in response.results:
            transcript += result.alternatives[0].transcript
    
        return transcript
    
    def generate_subtitles(input_video, transcript):
        clips = []
        video = VideoFileClip(input_video)
        video_duration = video.duration
        text_duration = 5  # 每行字幕的持续时间,单位为秒
    
        for i in range(0, len(transcript), text_duration):
            txt_clip = TextClip(
                transcript[i:i + text_duration],
                fontsize=24,
                color="white",
                size=video.size,
                method="caption",
                align="South",
                interline=10
            ).set_duration(text_duration).set_position(("bottom"))
    
            clips.append(txt_clip)
    
        subtitles = concatenate(clips, padding=-0.1)  # 设置间隔以处理重叠的字幕
        final_video = CompositeVideoClip([video, subtitles.set_start(0).set_duration(video_duration)])
    
        return final_video
    
    def main():
        input_video = "your_video.mp4"  # 视频输入路径
        outputFileWithoutAudio = "output_video_without_audio.mp4"  # 带有字幕但不带音频的输出视频路径
        output_audio = "output_audio.wav"  # 音频输出路径
    
        os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = "path/to/your/credentials.json"
    
        # 提取音频
        video = VideoFileClip(input_video)
        audio = video.audio
        audio.write_audiofile(output_audio)
    
        # 转录音频并生成字幕
        transcript = transcribe_audio(output_audio)
        final_video = generate_subtitles(input_video, transcript)
        final_video.write_videofile(outputFileWithoutAudio, codec="libx264", audio=False)
        final_video.close()
    
    if __name__ == "__main__":
        main()
    

    使用时,请将 your_video.mp4 替换为输入视频文件的路径,将 path/to/your/credentials.json 替换为 Google Cloud API 凭据文件的路径。 需要输出带有字幕但不带音频的视频,可以将 output_video_without_audio.mp4 替换为所需的输出路径。

    注意:此代码示例依赖于 Google Cloud API (付费服务)。要从相同的音频和视频生成带有音频的视频,可以在上述代码的基础上进一步修改 final_video.write_videofile() 方法中的 audio 参数。

  9. User

    写一段Python代码,根据视频内容编写一个脚本,并保存到文本文件

  10. Assistant

    为了根据视频内容编写脚本,我们需要从视频中提取音频,然后将音频转录为文本。本示例将使用 moviepy 提取音频并使用 Google Speech-to-Text API 将音频转录为文本。确保先安装所需的库:

    pip install moviepy
    pip install --upgrade google-cloud-speech
    

    接下来,将GOOGLE_APPLICATION_CREDENTIALS环境变量设置为 Google Cloud API 凭据的路径(JSON 文件),并运行以下代码:

    import os
    from google.cloud import speech_v1p1beta1 as speech
    from moviepy.editor import *
    
    def transcribe_audio(input_audio):
        client = speech.SpeechClient()
        audio = speech.RecognitionAudio(uri=input_audio)
    
        config = speech.RecognitionConfig(
            encoding=speech.RecognitionConfig.AudioEncoding.LINEAR16,
            language_code="en-US",
            enable_automatic_punctuation=True,
        )
    
        operation = client.long_running_recognize(config=config, audio=audio)
    
        response = operation.result(timeout=360)
    
        transcript = ""
        for result in response.results:
            transcript += result.alternatives[0].transcript
    
        return transcript
    
    def extract_audio_from_video(input_video, output_audio):
        video = VideoFileClip(input_video)
        audio = video.audio
        audio.write_audiofile(output_audio)
    
    def save_transcript_to_file(transcript, output_file):
        with open(output_file, "w") as f:
            f.write(transcript)
    
    def main():
        input_video = "your_video.mp4"  # 视频输入路径
        output_audio = "output_audio.wav"  # 音频输出路径
        transcript_output_file = "transcript.txt"  # 转录的输出文本文件路径
    
        os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = "path/to/your/credentials.json"
    
        # 从视频文件中提取音频
        extract_audio_from_video(input_video, output_audio)
    
        # 将音频文件转录为文本
        transcript = transcribe_audio(output_audio)
    
        # 将转录的文本保存到文件中
        save_transcript_to_file(transcript, transcript_output_file)
    
    if __name__ == "__main__":
        main()
    

    使用时,请将 your_video.mp4 替换为输入视频文件的路径,将 path/to/your/credentials.json 替换为 Google Cloud API 凭据文件的路径。根据需要,可以通过修改 transcript_output_file 变量来调整转录输出文本文件的路径。

    注意:此代码示例依赖于 Google Cloud API(付费服务)。此外,音频文件转录为纯文本,可能损失一些视频中的语境或信息。要以更多格式的文本保存,可以进一步修改 save_transcript_to_file() 函数。

  11. User

    根据视频提取音频,并生成字幕保存Srt文件

  12. Assistant

    为了从视频中提取音频并根据音频生成字幕(SRT 格式),我们将使用 moviepy 库提取音频并使用Google Speech-to-Text API 将音频转录为文本。然后,我们将使用 pysrt 库将转录的字幕转换为 SRT 格式。

    首先,安装必要的库:

    pip install moviepy
    pip install --upgrade google-cloud-speech
    pip install pysrt
    

    接下来,将 GOOGLE_APPLICATION_CREDENTIALS 环境变量设置为 Google Cloud API 凭据的路径(JSON 文件),确保API有效,并运行以下代码:

    import os
    import pysrt
    from google.cloud import speech_v1p1beta1 as speech
    from moviepy.editor import *
    
    def transcribe_audio(input_audio):
        client = speech.SpeechClient()
        audio = speech.RecognitionAudio(uri=input_audio)
    
        config = speech.RecognitionConfig(
            encoding=speech.RecognitionConfig.AudioEncoding.LINEAR16,
            language_code="en-US",
            enable_automatic_punctuation=True,
            enable_word_time_offsets=True
        )
    
        operation = client.long_running_recognize(config=config, audio=audio)
        response = operation.result(timeout=360)
    
        return response
    
    def extract_audio_from_video(input_video, output_audio):
        video = VideoFileClip(input_video)
        audio = video.audio
        audio.write_audiofile(output_audio)
    
    def generate_srt_file(response, output_file):
        subs = pysrt.SubRipFile()
        counter = 1
        for result in response.results:
            for word_info in result.alternatives[0].words:
                start_time = int(word_info.start_time.seconds * 1000 + word_info.start_time.nanos / 1e6)
                end_time = int(word_info.end_time.seconds * 1000 + word_info.end_time.nanos / 1e6)
                text = word_info.word
                subtitle = pysrt.SubRipItem(
                    index=counter,
                    start=pysrt.SubRipTime(millisecond=start_time),
                    end=pysrt.SubRipTime(millisecond=end_time),
                    text=text
                )
                subs.append(subtitle)
                counter += 1
    
        subs.save(output_file, encoding='utf-8')
    
    def main():
        input_video = "your_video.mp4"  # 视频输入路径
        output_audio = "output_audio.wav"  # 音频输出路径
        srt_output_file = "subtitles.srt"  # 字幕输出文件(SRT格式)
    
        os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = "path/to/your/credentials.json"
    
        # 从视频文件中提取音频
        extract_audio_from_video(input_video, output_audio)
    
        # 将音频文件转录为文本
        response = transcribe_audio(output_audio)
    
        # 将转录的文本保存到 SRT 文件中
        generate_srt_file(response, srt_output_file)
    
    if __name__ == "__main__":
        main()
    

    使用时,请将 your_video.mp4 替换为输入视频文件的路径,将 path/to/your/credentials.json 替换为 Google Cloud API 凭据文件的路径。根据需要,可以通过修改 srt_output_file 变量来调整字幕输出文件(SRT 格式)的路径。

    注意:此代码示例依赖于 Google Cloud API(付费服务)。音频转录为字幕后,可能丢失了部分字幕的上下文。要获取更准确的字幕,可能需要稍作编辑。

  13. User

    写一段Python代码,根据视频提取音频,并生成字幕保存Srt文件,不要用付费的Google Speech-to-Text API

  14. Assistant

    要从视频中提取音频并生成字幕(SRT 格式),我们可以使用 moviepy 库提取音频,并使用开源库 autosub 从中提取字幕。请先安装所需的库:

    pip install moviepy
    pip install autosub
    

    然后,运行以下代码:

    import os
    import autosub
    from moviepy.editor import *
    
    def extract_audio_from_video(input_video, output_audio):
        video = VideoFileClip(input_video)
        audio = video.audio
        audio.write_audiofile(output_audio)
    
    def generate_srt_file(input_audio, srt_output_file, language_code="en"):
        autosub.generate_subtitles(
            input_audio,
            src_language=language_code,
            output=srt_output_file,
            subtitle_file_format="srt"
        )
    
    def main():
        input_video = "your_video.mp4"  # 视频输入路径
        output_audio = "output_audio.wav"  # 音频输出路径
        srt_output_file = "subtitles.srt"  # 字幕输出文件(SRT 格式)
    
        # 从视频文件中提取音频
        extract_audio_from_video(input_video, output_audio)
    
        # 生成 SRT 字幕文件
        generate_srt_file(output_audio, srt_output_file)
    
    if __name__ == "__main__":
        main()
    

    使用时,请将 your_video.mp4 替换为输入视频文件的路径。根据需要,可以通过修改 srt_output_file 变量来调整字幕输出文件(SRT 格式)的路径。

    注意:autosub 库使用基于 Google Cloud API 的非官方 Speech-to-Text 实现,但不会产生费用,因为它使用免费的特定授权。速度可能较慢且不稳定。生成的字幕质量可能不如官方 Google Speech-to-Text API。 可以通过修改 generate_srt_file() 函数中的 language_code 参数来设置目标语言。默认为英文("en")。查看其他可用语言代码:https://cloud.google.com/speech-to-text/docs/languages